Our long range objective is to develop 3 dimensional radiotherapy methods (3DCRT) to increase the local control of primary tumors, based on the hypothesis that enhanced local control will result in improve survival due to (I) decreased metastatic spread, in addition to, (ii) decreased cancer death from local relapse. Recent advances have improved 3DCRT which may increase local control, to test the above hypothesis. In the first phase of this work, we improved our 3DCRT capability and began clinical studies. Phase I dose-escalation of prostate and lung is at 81 and 75.6 Gy, respectively. We have developed inverse planning optimization, and implemented intensity-modulated (IM) radiotherapy using dynamic multileaf collimation (DMLC). In parallel, we gathered data on normal tissue complications with 3D dose volume data, developed biophysical models, studied the treatment uncertainties from set-up and organ motion and assessed their effects on outcome. This renewal is a logical extension of the above. In Research Project I (RP I), with inverse planning and DMLC we shall continue Phase I study for prostate to 86.4 and 91.8 Gy to establish the maximum tolerable doses, and shall also begin a Phase II study for T1c/T2b patients at 81 Gy. For patients with lung carcinoma, we shall use a novel dose escalation protocol based on the combined criteria of dose and fractional organ damage calculated with a parallel model. In parallel, complication data will be studied in terms of biophysical models RP II concerns a generalized, improved method of optimization, to include both photons and electrons, non-coplanar fields, and IM beams, with versatile criteria, e.g. dose, dose-volume, and constraints of various forms. In RP III we shall develop methods to more accurately shape the field to the target. For the lung we shall minimize intrafraction tumor motion with the use of breath- hold, monitored by spirometry, and referenced to CT and fluoroscopic video. For the prostate, as interfraction uncertainty predominates, we shall (I) gather a population averaged data-base, and (ii) develop a statistical method of designing the intensity modulation beams. There aims are specific incremental steps towards our long term goals of increasing cancer cure by 3DCRT. Concomitantly, we shall improve the efficiency and the cost- effectiveness of 3DCRT, and effect technology transfer so as to broaden the use of 3DCRT.